Glu-66, Asp-138, and Asp-141, of RuvC constitute the catalytic center of the RuvC endonuclease. Molecular mechanisms of homologous DNA recombination.
Proc. Natl. Acad. Sci. USA Vol. 92, pp. 7470-7474, August 1995 Biochemistry
Identification of four acidic amino acids that constitute the catalytic center of the RuvC Holiday junction resolvase (homologous recombination/endonuclease/DNA repair/DNA binding)
ATSUSHI SAITO*, HIROSHI IWASAKI*, MARIKO ARIYOSHIt, KosuKE MORIKAWAt, AND HIDEO SHINAGAWA*t *Department of Molecular Microbiology, Research Institute for Microbial Diseases, Osaka University, Suita, Osaka 565, Japan; and Research Institute, Suita, Osaka 565, Japan
tProtein Engineering
Communicated by Kiyoshi Mizuuchi, National Institutes of Health, Bethesda, MD, April 27, 1995
Escherichia coli RuvC protein is a specific ABSTRACT endonuclease that resolves Holliday junctions during homologous recombination. Since the endonucleolytic activity of RuvC requires a divalent cation and since 3 or 4 acidic residues constitute the catalytic centers of several nucleases that require a divalent cation for the catalytic activity, we examined whether any of the acidic residues of RuvC were required for the nucleolytic activity. By site-directed mutagenesis, we constructed a series ofruvC mutant genes with similar amino acid replacements in 1 of the 13 acidic residues. Among them, the mutant genes with an alteration at Asp-7, Glu-66, Asp-138, or Asp-141 could not complement UV sensitivity of a ruvC deletion strain, and the multicopy mutant genes showed a dominant negative phenotype when introduced into a wildtype strain. The products of these mutant genes were purified and their biochemical properties were studied. All of them retained the ability to form a dimer and to bind specifically to a synthetic Holliday junction. However, they showed no, or extremely reduced, endonuclease activity specific for the junction. These 4 acidic residues, which are dispersed in the primary sequence, are located in close proximity at the bottom of the putative DNA binding cleft in the three-dimensional structure. From these results, we propose that these 4 acidic residues constitute the catalytic center for the Holliday junction resolvase and that some of them play a role in coordinating a divalent metal ion in the active center.
the sought-after enzyme responsible for Holliday junction resolution in E. coli (8-10). RuvC protein forms a stable homodimer of 19-kDa subunits, each 172 amino acid residues long (9). RuvC binds to a synthetic four-way junction with higher affinity than to normal duplex DNA (9, 10). This topology-specific binding does not require a divalent cation such as Mg2+ or Mn2+ (11, 12), which is required for the RuvC-catalyzed cleavage of Holliday junctions. To study the function-structure relationship of RuvC protein, we initiated mutational analysis of the RuvC function along with a crystallographic study. The crystallographic and mutational analysis of the 3'-5' exonuclease of E. coli DNA polymerase I (13), E. coli RNase Hi (14-17), and RNase H domain of human immunodeficiency virus type 1 reverse transcriptase (18-20) showed that 3 or 4 acidic residues constitute the catalytic centers of these enzymes and some of them are involved in coordinating divalent metal ion such as Mg2+. To examine whether any of the acidic residues in RuvC play an essential role in the resolvase activity, we used sitedirected mutagenesis to construct mutant ruvC genes with similar amino acid replacements in each of the 13 acidic residues and examined the function of the mutant genes in vivo and the properties of the mutant RuvC proteins in vitro. In this paper, we describe evidence that 4 acidic residues, Asp-7, Glu-66, Asp-138, and Asp-141, of RuvC constitute the catalytic center of the RuvC endonuclease.
Molecular mechanisms of homologous DNA recombination have been studied mainly in Escherichia coli and essential features of these processes appear to be conserved through the biological world (for reviews, see refs. 1 and 2). The processes of homologous recombination can be dissected into three steps. (i) Recombinogenic DNA molecules with a 3' singlestranded region to which RecA can bind to initiate recombination are formed by the combined helicase-nuclease functions of RecBCD, RecQ-RecN, or HelD-RecJ proteins. (ii) RecA protein, with the help of several recombination proteins, promotes homologous pairing of DNA molecules and initiates strand exchange between the paired molecules. The result is a recombination intermediate, called a Holliday junction (3), in which two homologous duplex DNA molecules are linked by a single-stranded crossover. In the subsequent steps, a RuvA and RuvB protein complex or RecG protein specifically binds to the Holliday junction and promotes branch migration by facilitating a strand-exchange reaction concomitant with ATP hydrolysis (4-7). The heteroduplex region is enlarged as a result of branch migration. To complete the recombination reaction and produce viable recombinant molecules, the Holliday junction has to be resolved by a specific endonuclease and subsequently sealed by DNA ligase. Recent genetic and biochemical studies have demonstrated that the RuvC protein is
MATERIALS AND METHODS tives of AB1157 (21), which was used as a ruv+ control strain. AruvC100::Cm and AruvC200::Km alleles were constructed on plasmids using the cloned gene, transferred into AB1157 via a recD strain (D301) as described (22). AruvC200::Km allele of HRS1200 was transferred into E. coli BL21 (DE3) by P1 transduction, generating HRS777. The mutant ruvC genes were constructed by site-directed mutagenesis using the cloned ruvC gene on M13mpl9 phage (23). The wild-type and mutant ruvC genes were excised from the phage vectors by cleaving at a Nco I site at the initiation codon and a BamHI site in the 3' flanking region, which were created by PCR-mediated sitedirected mutagenesis, and cloned into the Nco I and BamHI sites of pET-8c (24), a T7 expression plasmid. These plasmids were introduced into HRS1100 and AB1157 for complementation test. BL21(DE3) and HRS777 carrying pLysE (24) were used for overproduction of the wild-type and mutant RuvC proteins, respectively. RuvC Purification. The wild-type and mutant RuvC proteins were purified by the procedure slightly modified from the published one (9), which will be described elsewhere. The
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
*To whom reprint requests should be addressed at: Department of Molecular Microbiology, Research Institute for Microbial Diseases, Osaka University, 3-1, Yamadaoka, Suita, Osaka 565, Japan.
E. coli Strains and Plasmids. E. coli K-12 strains HRS1100
(AruvClOO::Cm) and HRS1200 (AruvC200::Km) are deriva-
7470
Biochemistry: Saito et al. concentration of RuvC was determined by using a molar extinction coefficient sM of 7.0 x 103 M-1-cm-1, which was obtained by a method of Gill and von Hippel (25). Synthetic Holliday Junction. The sequences of deoxyoligonucleotides used for making a model Holliday junction and a duplex DNA are as follows: HJ-1, 5'-GGTAGGACGGCCTCGCAATCGGCTTTGACCGAGCACGCGAGATGTCAACG-3'; HJ-2, 5'-CGTTGACATCTCGCGTGCTCGGTCAATCGGCAGATGCGGAGTGAAGTTCC-3'; HJ-3, 5'-GGAACTTCACTCCGCATCTGCCGATTCTGGCTGTGGCGTGTTTCTGGTGG-3'; HJ-4, 5'-CCACCAGA-
AACACGCCACAGCCAGAAAGCCGATTGCGAGGCCGTCCTACC-3'; D-5, 5 '-GGTAGGACGGCCTCGCAATCGGCTTTCTGGCTGTGGCGTGTTTCTGGTGG-3'. The model junction was constructed by annealing the four oligonucleotides HJ-1, HJ-2, HJ-3, and HJ-4. The underlined sequences constitute a 2-bp homology core in the central region of the four-way junction. Duplex DNA was made by annealing HJ-4 and D-5. HJ-4 was 5'-end-labeled with 32p prior to annealing. Holliday Junction Cleavage. The standard reaction mixture (20 gl) for cleavage of the synthetic junction, which contained RuvC (15 ng) and 32P-labeled substrate DNA (10 ng) in 20 mM Tris acetate, pH 8.0/10 mM magnesium acetate/i mM dithiothreitol/bovine serum albumin (100 tkg/ml)/5% (vol/vol) glycerol, was incubated for the indicated period at 37°C. Reactions were stopped by the addition of 5 gl of loading buffer (100 mM EDTA, pH 8.0/1% SDS/20% glycerol/0.1% bromophenol blue), and the reaction products were analyzed by PAGE on a 12% gel in TAE buffer (40 mM Tris acetate, pH 7.8/1 mM EDTA). The DNA bands were visualized and analyzed quantitatively by an image analyzer, Fuji BAS2000. Protein-DNA Binding Assay. The reaction mixtures (20 pkl) containing RuvC (20, 35, or 50 ng) and 32P-labeled substrate DNA (10 ng) in 20 mM Tris acetate, pH 8.0/0.5 mM magnesium acetate/i mM dithiothreitol/bovine serum albumin (100 ,tg/ml)/5% glycerol were incubated for 10 min on ice. To each sample, 5 pkl of loading buffer (20 mM Tris acetate, pH 8.0/10% glycerol/0.1% bromophenol blue) was added, and the samples were analyzed by PAGE in a 5% gel in TAM buffer (40 mM Tris acetate, pH 7.8/0.5 mM magnesium acetate) at 10 V/cm for 1 hr at 4°C. The DNA bands in the gels were visualized by the image analyzer.
RESULTS Identification of the Acidic Amino Acids Required for the RuvC Activity. To search for acidic residues in RuvC that are important for the catalytic activity, we made a series of the mutant ruvC genes to change each of seven Asp and six Glu residues of RuvC to Asn and Gln, respectively (Table 1). The wild-type and mutant alleles on pET-8c were introduced into HRS1100 and AB1157 to assay the recombination repair
Proc. Natl. Acad. Sci. USA 92 (1995)
7471
Table 1. Effects of ruvC mutations on DNA repair activity HRS1100 (AruvC) AB1157 (wild-type) ruvC % Relative % Relative Plasmid mutation survival activity, % survival activity, % pET-8c (vector) 0.17 1.4 11 100 12 pRC100 (ruvC+) 100 9.6 87 0.12 pRC101 D7N 1.0 0.14 1.3 0.21 pRC121 D7E 1.8 0.51 4.6 11 pRC102 D38N 92 9.5 86 17 pRC103 D39N 142 12 109 pRC104 D61N 11 92 16 145 pRC105 D75N 16 133 16 145 pRC106 D138N 0.17 1.4 0.21 1.9 pRC126 D138E 0.26 2.2 0.18 1.6 pRC107 D141N 0.14 1.2 0.15 1.4 pRC127 D141E 0.43 3.6 0.80 7.2 pRC141 E53Q 18 150 12 109 pRC142 E66Q 0.20 1.7 0.26 2.4 pRC162 E66D 0.16 1.3 0.18 1.6 pRC143 E95Q 16 133 13 118 pRC144 E1OOQ 23 192 15 136 pRC145 E117Q 19 158 12 109 pRC146 E161Q 20 167 14 127 Survival after UV-irradiation (40 J/m2) of exponentially growing cells carrying the indicated plasmids was measured. Average values of the three or four experiments are shown.
activity of the ruvC genes. Nine of the 13 mutant genes fully complemented the DNA repair deficiency of the ruvC deletion host, but 4 mutants with an alteration at Asp-7, Glu-66, Asp-138, or Asp-141 (D7N, E66Q, D138N, and D141N) did not. Furthermore, these four mutant genes on plasmids made the wild-type strain (AB1157) UV-sensitive despite the presence of the chromosomal ruvC+ allele (dominant negative mutations). We then constructed another series of the mutant ruvC genes to change these four acidic residues to the other acidic residues (D7E, E66D, D138E, and D141E). These mutant genes also did not complement the DNA repair deficiency of the ruvC deletion strain and made the wild-type strain UV-sensitive (Table 1). The levels of synthesis of these mutant RuvC proteins under the conditions of the above complementation test were examined by Western blot analysis using anti-RuvC serum (Fig. 1). In both the wild-type and AruvC hosts, all the mutant RuvC and RuvC+ proteins encoded by the plasmids were synthesized at almost the same level. Therefore, failure of complementation of the AruvC strain by the mutant genes was not due to the lack of expression or instability of the synthesized proteins. The amounts of the plasmid-coded RuvC proteins were 6- to 10-fold of that expressed from the chromosomal gene, as measured by densitometric scanning of the Western blot (Fig. 1). Therefore, the dominant negative effects of these mutant o
HRS1100
(AruvC) AB1157
(ruvC +)
4- RuvC
4- RuvC
FIG. 1. Expression of the RuvC proteins in the strains used for the complementation test. The plasmids containing the wild-type or mutant ruvC genes were introduced into strains HRS1100 (AruvC) and AB1157 (ruvC+). The cells were grown in LB medium containing ampicillin (50 ,ug/ml) at 37°C until the cell density reached an OD600 of 1.0 and were harvested. Total cell proteins were fractionated by SDS/PAGE in 12% gels and the RuvC proteins were detected by using rabbit anti-RuvC serum and an ECL Western blot kit (Amersham). WT, wild-type RuvC; (vector), cells with pET-8c.
7472
Biochemistry: Saito et al.
Proc. Natl. Acad. Sci. USA 92 (1995)
.41%
M
junction-
duplex-b
i
a b c d e f g h i
j k
FIG. 2. Cleavage of a model Holliday junction by the mutant RuvC proteins. The wild-type and mutant RuvC proteins were examined for the activity to cleave the junction in the standard reaction at 37°C for 10 min. The reaction products were separated by PAGE under nondenaturing conditions. -, Without RuvC; WT, wild type.
proteins in vivo are likely due to the inhibition of the function of wild-type RuvC protein by the mutant proteins expressed in the same cells. Asp-7, Glu-66, Asp-138, and Asp-141 of RuvC Play an Essential Role in the Cleavage of Holliday Junctions. To avoid even a trace of contamination of the wild-type RuvC expressed from the host chromosome to the mutant RuvC preparations, a ruvC deletion derivative of a BL21(DE3) was constructed and used as a host strain for overproduction of the eight mutant RuvCs (D7N, D7E, E66Q, E66D, D138N, D138E, D141N, and D141E). All of the mutant RuvCs were purified by the same procedure as RuvC+ to .99% purity (data not shown). All the mutant proteins gave very similar profiles at each chromatography step during purification as those of RuvC+. Therefore, these mutant RuvCs are likely to retain a tertiary structure very similar to that of RuvC+. To examine the ability of the mutant RuvCs to resolve Holliday junctions, we used a small synthetic four-way junction with a mobile central core as a model substrate for the Holliday junction. During a 10-min incubation under the standard conditions for the cleavage reaction (10 mM Mg2+), RuvC+ cleaved "80% of the junction DNA (Fig. 2, lane c). The dominant negative mutant RuvCs gave no or only a trace of cleaved products (Fig. 2, lanes d-k). The cleaved products made by the mutant RuvCs, E66Q and D141E, were detected only after overexposure of the gel or after prolonged reaction (Table 2). With respect to other mutant RuvCs, the cleaved product was not detected even after overexposure or prolonged incubation. These results suggest that the four acidic amino acids, Asp-7, Glu-66, Asp-138, and Asp-141, play a critical role in the nucleolytic reaction. We examined the ability of these mutant RuvCs to bind to Holliday junctions by a gel-retardation assay. The ability of all the mutant RuvCs to bind to the synthetic junction was indistinguishable from that of the wild-type RuvC (Fig. 3). This result shows that a defect in junction-specific binding was not WT o
1
Table 2. Effects of divalent cations at different concentrations on the cleavage activity of the mutant RuvC proteins % of substrate DNA cleaved Mn2+ Mg2+ RuvC 10 mM 20 mM 40 mM 10 mM 20 mM 40 mM 84 RuvC+ 87 92 78 80 83 D7N
